Table 8.

Effect of the probability of receiving short-time work compensation on firm behaviors at the cutoff of April sales change being below -5%.

(a) Passive measures
	(1)	(2)	(3)	(4)	(5)
	Business suspension	Employee suspension	Establishment closure	Underperforming Section closure	Cutting suppliers and buyers
Subsidy probability	0.772	0.145	0.278	0.958	0.0157
	(1.684)	(1.732)	(0.641)	(1.081)	(0.943)
Observations	3385	3385	3385	3385	3385
(b) Reactive measures
	(1)	(2)	(3)	(4)
	Facilitate IT	Work from home	Online sales	Develop new product and service
Subsidy probability	0.0488	1.454	0.133	0.346
	(1.319)	(2.647)	(1.014)	(0.946)
Observations	3385	3385	3385	3385

Notes: Standard errors are in parentheses. ${}^{***}$ for $p<0.01$, ${}^{**}$ for $p<0.05$, and ${}^{*}$ for $p<0.1$. Employment growth is log(expected employment at end-June 2020 + 1) subtracted by log(employment at end-2019). Positive investment is a binary indicator of more than JPY 10,000 planned investment from April to June 2020. Positive disinvestment is a binary indicator of planned disinvestment. Survival probability is the probability of continuing business until the end of 2020 (measured in %). Subsidy probability is the probability of receiving the short-time work compensation (measured in %). The bias is corrected and the standard errors are robust to heteroskedasticity.